Abstract

The maintenance of biodiversity crucially depends on the evolutionary potential of populations to adapt to environmental change. Accelerating climate change and extreme temperature events urge us to better understand and forecast evolutionary responses. Here, we harnessed the power of experimental evolution with the microbial model system yeast (Saccharomyces spp.) to measure the evolutionary potential of populations to adapt to future warming, in real- time and across the entire phylogenetic diversity of the genus. We tracked the evolution of thermal performance curves (TPCs) in populations of eight genetically and ecologically diverse species under gradually increasing temperature conditions, from 25 to 40 degrees C, for up to 600 generations. We found that evolving toward higher critical thermal limits generally came at a cost, causing a decrease in both thermal tolerance and maximum growth performance. The evolution of TPCs varied significantly between species with strong genotype- by- environment interactions, revealing two main trajectories: i) Warm- tolerant species showed an increase in both optimum growth temperature and thermal tolerance, consistent with the "hotter is wider" hypothesis. ii) Cold- tolerant species on the other hand evolved larger thermal breadth and higher thermal limits, but suffered from reduced maximum performance overall, consistent with the generalist or "a jack of all temperatures is a master of none" hypothesis. In addition, cold- tolerant species never reached the warm- tolerant species' upper thermal limits. Our results show that adaptive strategies to increasing temperatures are complex, highlighting the need to consider both within and between species diversity when predicting and managing the impacts of climate change on populations.